air cooling technology for power electronics thermal

Air Cooling Technology for Power Electronics Thermal Control

Desikan BharathanEmail: [email protected]

Phone: 303-887-4215Organization: National Renewable Energy Laboratory

Team members:Charlie King

Ken Kelly Sreekant Narumanchi

Project Duration: FY06 to FY09

DOE FreedomCAR and Vehicle Technologies ProgramAdvanced Power Electronics and

Electric Machines Projects FY08 Kickoff Meeting

National Transportation Research Center Knoxville, Tennessee

November 14, 2007

This presentation does not contain any proprietary or confidential information

NREL/PR-540-42340

DOE FreedomCAR APEEM Projects FY08 Kickoff Meeting

The Problem

• Eliminate liquid coolant loops.

• Enable heat rejection directly to the sink, namely, ambient air. Simplify.

• Maintain die temperature below specified operating limits for long-term reliability.

• Reduce cost and meet other FreedomCAR goals for year 2020 ($8/kW for a 55 kW traction system).


Description of Technology

Silicon chip

SolderDirect Bond CopperCu, AlN, Cu

TIM

Base plate

Pin finsAir

Q heat

Advantages

Air is the ultimate sink

Rejecting heat to air can eliminate intermediate fluid loops

Air is benign and need not be carried

Air is a dielectric, and can contact the chip directly

Drawbacks

Air has a low specific heat

Air is a poor heat-transfer fluid

Air density is low


Uniqueness of Project and Impacts

• Approach looks at simple thermal solutions.

• Offers viable heat rejection directly to the sink.

• Air cooling offers simplicity and least number of components.

• Air cooling offers high degree of reliability.


Accomplishments to DateAir-Cooling System for

Power E lectronics

A Schem atic D iagram

Air intake m anifold

A ir filter

A ir intake line

Centrifugal fan

Air distribution m anifold

M icro-fin heat exchanger (under chip assem bly PCB )

Am bient air at 30oC

Air exhaust to atm osphere


Accomplishments to Date

H

Wt

tb

Wb ~ 10 mmLb = 10 mm

FinFluid Passage

Hot Base

Plate

Micro-fin array geometryTypical values are:

W=130 μm

t = 65 μm

H = 13 mm and

tb = 1 mm


Accomplishments to Date

Goal


Accomplishments to DateModels show that:

• Air cooling can remove fluxes up to 150 W/cm2 for Silicon-based devices.

• Higher chip operating temperatures will increase the flux close to programmatic goal of 200 W/cm2.

Comparison with the use of an intermediate liquid cooling loop indicate that:

• Air cooling is simple, less costly, and reliable.

Test in Progress at NREL


Project Objectives for FY08• Implement air cooling

– Work with an industrial partner to accommodate technology on a working inverter module to meet FreedomCAR goals.

• Validate models and design approach.

• Contribute to advanced PE development cooling options– Work closely with ORNL

in development efforts to meet programmatic goals.

CFD model of air flow


Technical Approach for FY08a)

b)

c)

d)


Technical Approach for FY081. Design an air cooled micro-channel fin

heat exchanger for an inverter module

- in close collaboration with a chosen industrial manufacturer that meets the performance goals and other manufacturing constraints.

2. Incorporate air cooling device in a production model as a prototype.

3. Test performance of the module under realistic working conditions.

4. Validate design with test data.

5. Develop guidelines for performance estimation, cost, volume, weight and other measures for air cooling for use by the industry.

6. Develop second iteration design and demonstrate air-cooling as a viable option.

Flow model in split fins


Timeline

2007

Oct Nov Dec

2008

Jan Feb Mar Apr May Jun Jul Aug Sep

Develop next

generation design

Incorporate air cooling in a prototype

Develop design guidelines for the industry

Thermal testing/validation for micro-fin channel geometry at

lab scale

Consult with manufacturers for incorporating air cooling option

Initiate and fine tune design with constraints

Flow and thermal testing


The Challenges/Barriers

To be addressed:• Erosion concerns -- due to coolant impingement and

long term effects of jet impingement on copper fins.

• Potentially low blower efficiency.

• Potential noise generation due to air flow on fins.

• Difficulties with differential thermal expansions and resulting stresses.

• Convince industry this is a viable technology with advantages, while long-term reliability issues are addressed.


Beyond FY08 -- Integrate the developed thermal control technology with reliability tasks based on a systems level approach

• Analyze pulsed systems to power

• Generate a design tool for air cooling

- Provide industry with an easy-to-use tool to evaluate all aspects of performance and specifications.


Base plate

Copper micro-fins

Bottom enclosure

Channel

enclosure

air cooling technology for power electronics thermal

Documents